A pulse radiolytic study of the acid dissociation of ... - ACS Publications

Jan 26, 1987 - It is well-known that as a result of hydrogen bonding the pATa of 13.6 for acidic dissociation of the phenolic proton in salicyclic...
0 downloads 0 Views 559KB Size
J . Phys. Chem. 1987, 91, 4591-4595

4591

A Pulse Radiolytic Study of the Acid Dissociation of OH Protons in Radicals Related to Salicylic Acid' Q. Sun and Robert H. Schuler* Radiation Laboratory and Department of Chemistry, University of Notre Dame, Notre Dame, Indiana 46556 (Received: January 26, 1987; In Final Form: March 30, 1987)

The deprotonation of carboxylated benzosemiquinone radicals prepared by pulse radiolytic oxidation of dihydroxybenzoic acids has been examined by time-resolved absorption spectrophotometry. The pKa for dissociation of the OH proton in 3-carboxy-l,4-benzosemiquinoneis found to be 6.47 or 2.4 units higher than that in the unsubstituted radical. This difference presumably reflects internal hydrogen bonding by the carboxy group. This pK, is, however, well below that of the OH proton in salicyclic acid (13.6) so that hydrogen bonding is appreciably decreased by the delocalization of the unpaired spin in this radical. In the absence of proton acceptors equilibrationis relatively slow and in near neutral solutions is controlled by spontaneous loss of the proton from the acidic form with a half-period of 23 p s . Protonation of the basic form of the radical occurs at the diffusion-controlled rate. The rate constant for deprotonation by OH- is relatively low, 4.7 X lo7 M-' s-l , so that reaction with base becomes important only above pH 10. As a result this radical provides an excellent system for studying acid-base equilibration processes in near neutral solutions. Azide ion is shown to be an efficient catalyst which allows the acid-base equilibrium to be examined on the 10-ps time scale. Deprotonation is also catalyzed by the dihydroxybenzoic acid used as the radical source. Analogous studies on 4-carboxy-1,3-benzosemiquinonegive the pKa as 7.9. In spite of this high pK,, which indicates the rate constant for spontaneous dissociation of this radical to be >lo3 s-I, the rate constant for deprotonation by OH-, 4.9 X lo8 M-I s-', is considerably higher than in the case of 3-carboxy-l,4-benzosemiquinone.This latter comparison again indicates that the nature of the spin delocalization has an appreciable effect on the reaction kinetics.

It is well-known that as a result of hydrogen bonding the pKa of 13.6 for acidic dissociation of the phenolic proton in salicyclic acid (2-hydroxybenzoic acid) is considerably higher than that in phenol (9.9) or in its isomers, 3-hydroxybenzoic acid (9.9) and 4-hydroxybenzoic acid (9.3).* During the course of pulse radiolytic studies of the semiquinone radicals produced in the oxidation of aqueous solutions of 2,4- and 2,s-dihydroxybenzoic acid we have observed that the pKas for acidic dissociation of their OH protons (6.47 and 7.9, respectively) are several units higher than those for the corresponding unsubstituted semiquinones, reflecting internal hydrogen bonding in these radicals similar to that in salicylic acid. Because of this hydrogen bonding, oxidation of these dihydroxybenzoic acids in acidic solutions leads preferentially, ~ ,the ~ formation of the isomer of as is shown by ESR s t u d i e ~ ,to the semiquinone with the salicyclic acid group intact, i.e. oxidation of 2,s-dihydroxybenzoic acid yields only 3-carboxy- 1,Cbenzosemiquinone 3-carboxy-4-hydroxyphenoxylradical). In effect H

0

H

0

QH'

the second OH group in these substrates provides a site for oxidative attack by which intermediates suitable for time-resolved studies of the dissociation of weak acids can be rapidly prepared. Hyperfine splitting by the OH proton of these radicals is observed in solutions near to neutral, showing that proton exchange with water is relatively slow, Le. that dissociation of the acidic form (1) This is Document No. NDRL-2926 from the Notre Dame Radiation Laboratory. (2) Serjeant, E. P.; Dempsey, B. Ionization Constants of Organic Acids in Aqueous Solution; Pergamon: Oxford, 1979. ( 3 ) Neta, P.; Fessenden, R. W. J. Phys. Chem. 1974, 78, 523. (4) Jinot, C.; Madden, K. P.; Schuler, R. H. J . Phys. Chem. 1986, 90, 4979.

cannot occur on a time scale much shorter than microseconds. In strongly basic solutions these semiquinones deprotonate by reaction with base to form the relatively unreactive carboxylated semiquinone radical dianions, e.g. QH'-

+ OH-

k-3

+

Qo2- H 2 0

(3)

We show here that the rate constants for deprotonation of these carboxylated benzosemiquinones by base are several orders of magnitude less than for diffusion-controlled reactions, making it possible to examine the other factors which control deprotonation in near neutral solutions. It is shown, for example, that the azide ion is an efficient proton acceptor by which one can catalyze the otherwise slow equilibration between the acidic and basic forms of the radical. The rate constant for spontaneous dissociation of the OH proton in 3-carboxy-l,4-benzosemiquinone( k 2 )is found to be 3 X lo4 s-l, indicating that k-2 is 8.5 X 10'O M-' s-l, i.e. that protonation of the radical dianion is essentially controlled by diffusion.

Experimental Section Time-resolved spectrophotometric studies were carried out by pulse radiolysis methods similar to those used in many recent investigations from these laboratorie~.~Most studies used N3' radicals produced by *OH oxidation of sodium azide (Fluka) as a secondary oxidant.6 In certain cases Br2*- produced from potassium bromide (Fisher) was used. Formation of these oxidizing radicals at 0.1 M azide or bromide was complete within (5) For a detailed description of the pulse radiolysis apparatus see: Janata, E.; Schuler, R. H. J . Phys. Chem. 1982.86, 2078. A Biomation 8100 was used for digitization of the transient signals in the present experiments. (6) Alfassi, Z. B.; Schuler, R. H. J . Phys. Chem. 1985, 89, 3359.

0022-3654/87/2091-4591$01.50/0 0 1987 American Chemical Society

4592 The Journal of Physical Chemistry, Vol. 91, No. 17, 1987

Sun and Schuler

TABLE I rate constant for oxidation, M-1 s-I x 10-9

substrate pH 2,5-dihydroxybenzoic -4.5 acid 11.6 2,4-dihydroxybenzoic -5.5 acid 10.5

-

forma monoanion dianion monoanion dianion

Ngob 2.6 4.2

Br2'-C

0.065

,p"

0.77

1 .O

0.055

3.5

0.72

pK,= 0 . 3

-

"The carboxy group of these benzoic acids is fully ionized above pH 5 (+K, 3). pK,, for ionization of the OH proton at C5 in 2,5-dihydroxybenzoic acid is 10.3 and that at C4 in 2,4-dihydroxybenzoic acid is 8.9 (see text). bFrom the pseudo-first-orderperiod for production of product radical. CFromthe pseudo-first-orderperiod for loss of Br2*- at an ionic strength of 0.1.

-

10 ns of the pulse. The dihydroxybenzoic acids (Aldrich) were present at 0.5 to 5 m M so that there was very little direct oxidation of the substrate by 'OH. This approach largely avoids the intermediary production of hydroxycyclohexadienyl radicals which otherwise complicates the reaction scheme where *OH is used directly.6 Studies were carried out in aqueous solutions which were purged of oxygen and saturated with NzO to convert eaqto 'OH. Radical concentrations were -5 pM. The pH was adjusted with HClO, or KOH and determined with an Orion 801 pH meter calibrated with Fisher buffers. Because of possible ionic strength effects care was used not to introduce extraneous ions so that the rate data are, in most cases, representative of an ionic strength of 0.1.

i

8

7

10

9

12

II

I3

PH

Figure 1. Dependence of the rate constant for oxidation of 2,5-dihydroxybenzoic acid by Br2'- on pH. The inflection point corresponds to a pK,, of 10.3 for the OH proton at position 5. Solid curve is calculated from this pK, and the limiting rate constants given in Table I. 1500

r

6000

4500

'E 'x ,

3000

c z 0 Y

Results and Discussion

1500

-

Rate constants for oxidation of 2,4-dihydroxybenzoic acid and 2,5-dihydroxybenzoic acid by N3' and Brz'- a t pH -5 and 1 1 are given in Table I. Oxidation by these secondary radicals takes place by electron transfer from the substrate to the oxidant? There is no apparent delay in formation of the semiquinone radicals, oxeven in acidic solutions, so that, as with 1,2,4-ben~enetriol,~ idation of the monoanion must be rapidly followed by proton loss. In the case of N3*,rate constants were determined from the pseudo-first-order period for growth of the absorption of the product radical and in the case of Br;- from the period for decay of the latter as measured at 360 nm. The carboxy group of both substrates is somewhat more acidic than that in benzoic acid (pKa, = 4.2 for benzoic acid, 3.0 for 2,5-dihydroxybenzoic acid, and 3.3 for 2,4-dihydroxybenzoic acid)2 so that it should be fully ionized at pH 5 and above. The second pKa of 2,4-dihydroxybenzoic acid (given in the literature as 9.0)2 is similar to the lowest value for resorcinol (9.3) and presumably represents ionization of the OH proton at C4. We have determined this pKa from the pH dependence of the rate constant for oxidation by Br2*- and find a value of 8.9, in good agreement with the literature value. The third pK, (- 14)2is very high and from its similarity to pKa2for salicylic acid is attributed to ionization of the OH adjacent to the carboxy group. The second and third pK,s of 2,5-dihydroxybenzoic acid are not available in the literature. We have examined its absorption spectrum and find a shift from a maxmum at 320 nm in neutral solution to one at 346 nm in base. The pH dependence indicates pKa2to be 10.4. Because the rate constant for oxidation by Brz*- increases by an order of magnitude on ionization of the OH (seeTable I) this oxidation rate also provides a good measure of the relative contribution of the acidic and basic forms. As indicated in Figure 1 the rate constant follows a titration curve very well with an inflection point corresponding to pKa2= 10.30 f 0.05 at an ionic strength of 0.1. This value is of the magnitude of the lowest pK, of hydroquinone (9.9) and is, therefore, attributed to ionization of the OH proton at Cs. The carboxy group clearly has a substantial effect on the acidicity of the remote OH proton, increasing its pKa in the substituted hydroquinone by 0.4 and decreasing its pKa in the substituted resorcinol by 0.3. The third

-

(7) Qin, L.; Tripathi, G. N. R.; Schuler, R. H. J. Phys. Chem., in press.

8 0 c

5

3000

w x

1500

0

350

400

450

WAVELENGTH

-

500

550

nm

Figure 2. Absorption spectra of the carboxysemiquinone radicals and their anions observed in the pulse radiolytic oxidation of 1 mM solvents of (A) 2,5-dihydroxybenzoic acid and (B) 2,4-dihydroxybenzoic acid. Spectra were recorded 10 ps after the pulse. Solid symbols are for studies at pH - 5 and open symbols at pH -10. Azide radical was used as a secondary oxidant in these studies so that the spectra are free of contributions from H atom addition products (see ref 8).

pKa of 2,s-dihydroxybenzoic acid, attributable to the OH proton at Cz, is unavailable but is expected to be similar to that for salicylic acid, Le. in the region of 13-14. One, therefore, expects that, as indicated by eq 1, oxidation of these substrates will initially yield semiquinone radicals with the oxygen atom adjacent to the carboxy group protonated even up to high pH. The rate constant for oxidation of the dianionic forms of these substrates by N3*is comparable to that for oxidation of most phenoxides (Le., (4-5) X IO9 M-' s-l)? Oxidation by N,' in acidic solutions is also comparable to that for oxidation of hydroquinone (4.2 X lo9 M-ls-' ) a nd resorcinol (1 .O X 1O9 M-I s-l ) a nd considerably faster than for oxidation of phenol (5 X lo7 M-I sbl Oxidation of these substrates by Brz'- is slower, particularly in acidic solutions where the oxygen atoms are fully protonated. Where it can be used, N3*is clearly the better choice as a secondary oxidant for preparing the semiquinone radicals. Absorption spectra of 3-carboxy- 1,4-benzosemiquinone and 4-carboxy- 1,3-benzosemiquinone (4-carboxy-3-hydroxyphenoxyl radical) observed at pH -5 (open points) and their corresponding semiquinone radical dianions observed at pH 11 (closed points) are illustrated in Figure 2. These spectra were recorded 10 I.LS after pulse irradiating solutions 0.1 M in azide and 1 mM in substrate. From the kinetic data presented below oxidation of

-

-

The Journal of Physical Chemistry, Vol. 91. No. 17, 1987 4593

Deprotonation of Carboxylated Benzosemiquinones

n

U

0

I

4

5

6

8

7

9

10

11

PH

I ,

Figure 3. Fraction of radicals present as dianion: ( 0 )for 3-carboxy1,4-benzosemiquinoneas determined at 432 nm and (0)at 460 nm; (A)

for 4-carboxy-1,3-benzosemiquinoneas determined at 455 nm. Calculated titration curves correspond to pK,s of 6.47 and 7.9, respectively.

the substrates should be complete and equilibrium established between the acidic and basic forms of the radical at the time of these recordings. Because H atoms are removed by reaction with azide ions these spectra are free of contributions from H atom addition products. Extinction coefficients are based on radical yields of 6.1 in acidic solution and 5.5 in basic solution.s The absorption spectrum of the 2-carboxy- 1,Cbenzosemiquinone radical dianion is very similar to that of the 1,Cbenzosemiquinone radical anion.9 Maxima are observed at 408 and 432 nm which are shifted to the red from those in the unsubstituted radical by -5 nm. Both Ramanlo and ESR3 data show that the carboxy substituent has relatively little effect on the electron distribution over the 7r system of this radical. In contrast the carboxylated benzosemiquinone observed on protonation exhibits a rather different spectrum than that of either phenoxy" or 1,4-benzosemiquinone radicaL9 In particular, there is appreciable absorption at wavelengths longer than 430 nm. For present purposes the relatively high absorption in the region of 460 nm is important in that it allows one to follow the decay of the protonated radical as well as the growth of the complementary semiquinone anion. The existence of an isosbestic point at 443 nm also allows one to examine the time dependence of the total radical population directly. The spectrum of 4-carboxy- 1,3-benzosemiquinone radical dianion is very similar in structure and intensity to that of 1,3benzosemiquinone radical anion.12 ESR datal3 show that the unpaired spin distribution in this radical is essentially unaffected by the presence of the carboxy group. The spectrum of 4carboxy- 1,3-benzosemiquinoneobserved on protonation at pH 5 is very much weaker and even less intense than that of the 1,3benzosemiquinone radical.12 As a result there is no isosbestic point and for kinetic purposes one can only follow growth of the completely ionized radical. The acidic dissociation constants of these radicals were detemined by examining the pH dependence of the absorption spectra. Measurements were made in 0.1 M sodium azide solution at 15 ps, at which time acid-base equilibrium is established (see below). The data are given in Figure 3 in terms of the fractional contribution of the basic form (the radical dianion) as determined from the intensities observed both at 432 and 460 nm in the case of the 1,4-semiquinone and at 455 nm in the case of the 1,3semiquinone. In the first instance the total radical yield could be monitored at the 443-nm isosbestic point, allowing corrections to be made for the slight changes in yield that occur in the pH range of 5 to 7.8 The pK,s observed for dissociation of the O H

I

4 I

II

0 380

400

I

I

420

440

,

I

1

460

480

WAVELENGTH

Figure 4. Time dependence of the spectrum (0.2, 0.6, 1.2, 2.0, 3.0, 4.4,

and 13 ps) in the oxidation of 1 mM 2,5-dihydroxybenzoicacid with N3'. Isobestic points are at 396,402, and 443 nm as indicated by the dashed lines. Spectra at 0.2 and 0.6 ps are corrected for incomplete oxidation as observed at the 443-nm isobestic point. Arrows indicate directions of absorbance changes with time.

moor

I1

A#-

4000

1:

m

a

o

b

1

0

05

10

15

10

20

30

W

Lc a 8000

J

w U

t

*.

6000 -

B

-

(8) Ye, M.; Madden, K. P.; Fessenden, R. W.; Schuler, R. H. J . Phys. Chem. 1986, 90, 5391. (9) A d a m , G . E.; Michael, B. D. Trans. Faraday SOC.1967, 63, 1171. (!O) Sun, Q.; Tripathi, G. N. R.; Schuler, R. H., to be submitted for

publication.

(11) Tripathi, G . N. R.; Schuler, R. H. J . Chem. Phys. 1984, 82, 113.

(12) Tripathi, G. N. R.; Chipman, D. C.; Miderski, C. A.; Davis, H. F.; Fessenden, R. W.; Schuler, R. H. J . Phys. Chem. 1986, 90, 3968. (13) Madden, K. P.; McManus, H. J.; Schuler, R. H.J. Phys. Chem. 1982,

86, 2926.

--

0

TIME-

/LS

Figure 5. Time dependence of absorbance in the oxidation of 5 mM 2,5-dihydroxybenzoic acid by N,' in the presence of 0.1 M N3- at pH 11.1: at ( 0 )543 nm, (0) 443 nm, and (A)460 nm. The half-period for

oxidation as measured at 432 nm is 34 ns and for subsequent deprotonation as measured at 431 and 460 nm is 1.80 pus.

*

protons at an ionic strength of 0.1 are 6.47 0.05 for 3carboxy- 1,4-benzosemiquinone and 7.9 0.1 for 4-hydroxy- 1,3benzosemiquinone. Similar pH dependences were observed for the radicals prepared by Br2*-oxidation but because of the low oxidation rates and relatively longer equilibration periods the inflection points (pKas of 6.5 and 8.1, respectively) have substantial possible errors. The observed pKas are several units higher than those for their unsubstituted counterparts (4.0 for 1,4-benzosemiquinoneg and 6.43 for 1,3-ben~osemiquinone'~).It is clear

*

(14) Jinot, C.; Madden, K. P.; Schuler, R. H. J . Phys. Chem. 1986, 90, 4919.

Sun and Schuler

The Journal of Physical Chemistry, Vol. 91, No. 17, 1987

4594

60

40

w;

0

c :2 2 U

w

z

Y

+

20

L

P 0

0

+ v)

L

00

020

0 10

0

::

[Nil

0 30

-M

Figure 7. Dependence of first-order rate constant on N3-concentration

v) Y

( 0 )3-carboxy-l,4-benzosemiquinone *from Figure 6 ) and (0) 4-carboxy-l,3-benzosemiquinone.Slopes correspond, respectively,

at pH -9:

to rate constant of 2.0

,-

lo6 and 1.8

X

X

lo6 M-I s-' for reaction 4.

9000 W

0

V

00

IO

2 0

[OH']

3 0

x

4 0

50

4

m

g

IO'-M

6000

v)

Figure 6. Dependence of measured first-order deprotonation rate con-

stant on concentration of base for solutions 0.1 M in N,:

m

4

( 0 ) 3-

carboxy-l,4-benzosemiquinone;(0)4-carboxy-l,3-benzosemiquinone. Slopes of lines correspond, respectively, to rate constants of 4.9 X lo' and 4.9 X lo8 M-I s-l for reaction 3.

that hydrogen bonding makes the O H protons in these radicals appreciably less acidic than in the unsubstituted benzosemiquinones. They are, however, much more acidic than the O H proton of salicylic acid (pK, = 13.6). The spectrum initially observed on oxidation of a 5 mM solution of 2-carboxy-5-hydroxybenzoicacid by N3*at pH 11 is identical with that found in acidic solutions, assuring that, as expected, the semiquinone radical is initially produced. Figure 4 illustrates the time dependence of the spectrum, as observed at pH 9.6. These spectra exhibit isosbestic points at 396, 402, and 443 nm which indicate that deprotonation occurs without complication. These isosbestic points also make it possible to compare the intensities of the acidic and basic forms of the radical directly without any assumptions as to the pH dependence of the yields. Typical kinetic traces are given in Figure 5 for a 0.1 M N3solution 5 mM in 2,5-dihydroxybenzoic acid at pH 11.1, It is seen from the trace at the 443-nm isosbestic point (curve 4b) that oxidation is for the most part complete in 200 ns. At longer times the absorption at 432 nm increases and that at 460 nm decreases with measured half-periods of 1.8 1 and 1.79 RS, respectively. The first-order rate constant for deprotonation under the conditions for this experiment is, therefore, 3.85 X lo5 s-I. As indicated in Figure 6 similar rate measurements on solutions 0.1 M in sodium azide show a linear dependence of the first-order rate constant on OH- concentration with a slope corresponding to a second-order rate constant for deprotonation by base of 4.9 X lo7 M-' s-l, The intercept of Figure 6 should be less than 3.5 X 104 s-' if the rate in neutral solutions is controlled by spontaneous deprotonation. This upper limit is estimated from the pK, (K2 = 3.5 X lo-' M) by assuming a maximum value of 10" M-' s-' for k-2. The pronounced intercept in Figure 6 (3.6 X lo5 s-I) is, however, an order of magnitude greater, implying that deprotonation occurs by some additional mechanism in solutions near to neutral. To a large extent this intercept can be attributed to deprotonation by azide ion, i.e.

-

QH'-

+ N3- eQo2-+ HN3 k4

(4)

6

I

I

The Journal of Physical Chemistry, Vol. 91. No. 17, 1987 4595

Deprotonation of Carboxylated Benzosemiquinones

b"

40r

"t

5 2.0 a

protons at a diffusion-controlled rate. Measurements similar to the above were also carried out on 2,4-dihydroxybenzoic acid and the rate data obtained are included in Figures 6 and 7. It is seen in Figure 6 that the rate constant for deprotonation of the 1,3-benzosemiquinoneby OH- is relatively high (4.9 X lo8 M-' s-l ). Deprotonation by N3- (Figure 7; k4 = 1.8 X lo6 M-' s-l ) is comparable in rate to that for its 1,4-isomer. The intercept noted for the this plot in Figure 7, 4.5 X IO4 s-I, appears to be almost entirely attributable to reaction 5 which is more important here than for 3-carboxy- 1,4-benzosemiquinone because of the close match of the pKs of the radical and the substrate. In this latter case, the relatively high pK, of the radical indicates that the rate constant for spontaneous dissociation must be less than lo3 SKI, Le. one expects the intrinsic lifetime of the protonated radical to approach 1 ms. However, the maximum lifetime that we were able to demonstrate was only 60 ~s in an experiment 0.5 mM in 2,4-dihydroxybenzoic acid at pH 9.5. Under these conditions deprotonation is controlled by contributions of - 5 X IO3 s-l from both reactions 3 and 5 . Basically, it was not possible to design a practical experiment to determine the rate for spontaneous decay of this radical with any accuracy and we could only show that k2 cannot be significantly greater than the upper limit of lo3 s-l estimated from the equilibrium constant. The various kinetic data are summarized in Table 11. The carboxylated 1,4-benzosemiquinone produced on oxidation of 2,5-dihydroxybenzoic acid provides a particularly good system for studying the factors influencing the dissociation of an OH proton in the salicylic acid configuration. In this case the ionization equilibrium is controlled by dissociation on the 10-ps time scale and protonation at a diffusion-controlled rate. Ionization of 4-carboxy- 1,3-benzosemiquinone is somewhat slower. Because of the relatively low rates intrinsic to these reactions, catalysis by even modest concentrations of donors and acceptors becomes important. Internal hydrogen bonding is important in controlling the ionization, as is illustrated by the relatively high pK,s and low rates for deprotonation by base. Comparison of the data on these two radicals shows that the delocalization of the unpaired spin in 3-carboxy- 1,Cbenzosemiquinone increases the dissociation rate considerably over that for its 1,3-benzosemiquinone isomer.

ii I-

v)

[r LL

00

00

02

DIANION

04

CONCENTRATION

06

103-~

Figure 9. Dependence of the rate constant for deprotonation of 3carboxy-1,4-benzosemiquinoneon the concentration of the dianion of 2,5-dihydroxybenzoic acid: (0)Experiments at p H -8.7 where -3% of the substrate is in the dianionic form, (a) at pH 9.5 where -15% is in the dianionic form. TABLE II: Equilibrium and Rate Constants for Deprotonation

K2equilibrium, M-I k2 spontaneous dissociation, s-I k-, protonation, M-' s-I k, deprotonation by base, M-l s-I k4 deprotonation by N), M-I s-I k5 deprotonation by substrate! M-I s-1

3-carboxy-1,44-carboxy-1,3benzosemiquinone benzosemiquinone 3.4 x 10-7 1.2 x 10-8 3.4 x 104